Human Thoracolumbar Spine Tolerance to Injury and Mechanisms From Caudo-Cephalad Impacts

Objective: Injury criteria development continues to be a research topic for the lumbar spine, for which standards do not exist.  As axial loading of the lumbar spine occurs in automotive (frontal impact to restrained occupants) and military events, it is important to determine tolerance in this mode. The aims of this study are to determine injury risk curves using post mortem human surrogate (PMHS) specimens.

Methods: T12-S1 (T12-L1 and L5-S1 joints unconstrained) PMHS lumbar columns were subjected to impact along the caudo-cephalad direction (velocities: 0.3-12.3 m/s). Forces at caudal and cephalad ends were recorded using load cells and transformed to most proximal disc centers in the time domain. Pretest quantitative computed tomography images were used to extract vertebral level-specific disc and body areas, and column lengths were obtained. All these variables and age were used in risk curve analysis. Repeated testing protocols were used to bracket sub-injurious and injurious impacts. Radiographs, palpation, and comparison with previous history of loading and status were used to ensure the appropriateness for subsequent higher impact severity of. Further loading was ceased upon suspicion of injury, soft or hard tissue or excessive joint laxity. Injuries were categorized three different ways: AIS, stable versus unstable, and single versus multilevel. Peak axial and shear forces at the ends were used as primary response variables with other stated parameters as covariates in the parametric survival analysis. Brier scores was used to identify the best metric that explained the underlying injury mechanism.

Data Sources: Mean age, stature, mass, body mass index, column height, and disc and body areas of 40 injured and three intact columns: 65±10 years, 1.78±0.06 m, 83±13 kg, 26±4 kg/sqm, 17.8±0.9 cm, 18.9±2.8 sq cm, and 12.4±1.6 sq cm.

Results: Injuries consisted of compression fractures with/without disc, and with/without posterior complex involvements. Number of injuries were one-half in the single and multilevel, and stable and unstable groups. The same specimens were not in both groups. Resultant forces at the caudal end for stable and resultant forces at the cepahald end for the unstable injuries were the optimal according to the Brier Score metric. Peak forces of 5.8kN and 7.6kN were associated with 50% probabilities for stable and unstable injuries, and the ±95% confidence intervals (CI) were such that the two curves were considered to have a good quality indices. A similar result was found for single and multilevel injuries.  Individual and covariate-based risk curves will be provided if the abstract is accepted. 

Significance of results: This is the first study to investigate and replicate injuries to lumbar spine from caudo-cepahald impacts and develop injury risk curves using a host of biomechanical variables, conduct experiments in the same laboratory setting without confounding effects of device-related issues, and use largest dataset.  Results and data from this investigation can be used in dummies and computational models including isolated component spine and total human body models for a better understanding of the biomechanics of load transfer within the spinal components and injury.